Update app.py

app.py

@@ -1,486 +1,155 @@
-import streamlit as st
-import numpy as np
-import pandas as pd
-import matplotlib.pyplot as plt
-import seaborn as sns
+import gradio as gr
+import os
 import requests
-
-# Set page config
-st.set_page_config(
-    page_title="Heart Disease Prediction",
-    page_icon="❤️",
-    layout="wide"
-)
-
-# Mistral API Key
-MISTRAL_API_KEY = "yQdfM8MLbX9uhInQ7id4iUTwN4h4pDLX"
-
-# Feature information with medical ranges
-feature_info = {
-    "age": {"label": "Age", "desc": "Age in years", "min": 20, "max": 100, "value": 50,
-            "healthy_range": "20-60", "risk_range": "60+"},
-    "sex": {"label": "Sex", "desc": "Biological sex", "options": ["Female", "Male"]},
-    "cp": {"label": "Chest Pain Type", "desc": "Type of chest pain",
-           "options": ["Typical angina", "Atypical angina", "Non-anginal pain", "Asymptomatic"],
-           "healthy": "Typical angina", "risk": "Asymptomatic"},
-    "trestbps": {"label": "Resting BP", "desc": "Resting blood pressure (mm Hg)",
-                 "min": 80, "max": 200, "value": 120,
-                 "healthy_range": "90-120", "risk_range": "140+"},
-    "chol": {"label": "Cholesterol", "desc": "Serum cholesterol (mg/dl)",
-             "min": 100, "max": 600, "value": 200,
-             "healthy_range": "< 200", "risk_range": "240+"},
-    "fbs": {"label": "Fasting Blood Sugar", "desc": "> 120 mg/dl", "options": ["No", "Yes"],
-            "healthy": "No", "risk": "Yes"},
-    "restecg": {"label": "Resting ECG", "desc": "Resting electrocardiographic results",
-                "options": ["Normal", "ST-T wave abnormality", "Left ventricular hypertrophy"],
-                "healthy": "Normal", "risk": "Left ventricular hypertrophy"},
-    "thalach": {"label": "Max Heart Rate", "desc": "Maximum heart rate achieved",
-                "min": 60, "max": 220, "value": 150,
-                "healthy_range": "60-100 (resting)", "risk_range": "< 120 (exercise)"},
-    "exang": {"label": "Exercise Angina", "desc": "Exercise induced angina", "options": ["No", "Yes"],
-              "healthy": "No", "risk": "Yes"},
-    "oldpeak": {"label": "ST Depression", "desc": "ST depression induced by exercise",
-                "min": 0.0, "max": 6.0, "value": 1.0, "step": 0.1,
-                "healthy_range": "0-1", "risk_range": "2+"},
-    "slope": {"label": "ST Slope", "desc": "Slope of peak exercise ST segment",
-              "options": ["Upsloping", "Flat", "Downsloping"],
-              "healthy": "Upsloping", "risk": "Downsloping"},
-    "ca": {"label": "Major Vessels", "desc": "Number of major vessels colored by fluoroscopy",
-           "options": ["0", "1", "2", "3"],
-           "healthy": "0", "risk": "3"},
-    "thal": {"label": "Thalassemia", "desc": "Blood disorder called thalassemia",
-             "options": ["Normal", "Fixed defect", "Reversible defect"],
-             "healthy": "Normal", "risk": "Fixed defect"}
-}
-
-# Risk assessment thresholds
-RISK_THRESHOLDS = {
-    "Low": 0.3,
-    "Moderate": 0.6,
-    "High": 1.0
-}
-
-# Sample healthy and unhealthy profiles
-SAMPLE_PROFILES = {
-    "Healthy": {
-        "age": 45, "sex": "Female", "cp": "Typical angina",
-        "trestbps": 110, "chol": 180, "fbs": "No",
-        "restecg": "Normal", "thalach": 160, "exang": "No",
-        "oldpeak": 0.5, "slope": "Upsloping", "ca": "0",
-        "thal": "Normal"
+from datetime import datetime
+
+# Mistral API setup
+MISTRAL_API_KEY = os.getenv("MISTRAL_API_KEY", "your_api_key_here")
+MISTRAL_API_URL = "https://api.mistral.ai/v1/chat/completions"
+
+# Local medical knowledge base
+MEDICAL_KNOWLEDGE = {
+    "headache": {
+        "assessment": "Could be tension-type, migraine, or other common headache",
+        "recommendations": [
+            "Rest in a quiet, dark room",
+            "Apply cold compress to forehead",
+            "Hydrate well (at least 8 glasses water/day)",
+            "Try gentle neck stretches",
+            "Consider OTC pain relievers (ibuprofen 200-400mg or paracetamol 500mg)"
+        ],
+        "warnings": [
+            "If sudden/severe 'thunderclap' headache",
+            "If with fever, stiff neck, or confusion",
+            "If following head injury"
+        ]
     },
-    "Moderate Risk": {
-        "age": 58, "sex": "Male", "cp": "Atypical angina",
-        "trestbps": 135, "chol": 230, "fbs": "No",
-        "restecg": "ST-T wave abnormality", "thalach": 140, "exang": "No",
-        "oldpeak": 1.5, "slope": "Flat", "ca": "1",
-        "thal": "Reversible defect"
+    "fever": {
+        "assessment": "Common symptom of infection or inflammation",
+        "recommendations": [
+            "Rest and increase fluid intake",
+            "Use light clothing and keep room cool",
+            "Sponge with lukewarm water if uncomfortable",
+            "Paracetamol 500mg every 6 hours (max 4 doses/day)",
+            "Monitor temperature every 4 hours"
+        ],
+        "warnings": [
+            "Fever >39°C lasting >3 days",
+            "If with rash, difficulty breathing, or seizures",
+            "In infants <3 months with any fever"
+        ]
     },
-    "High Risk": {
-        "age": 65, "sex": "Male", "cp": "Asymptomatic",
-        "trestbps": 180, "chol": 280, "fbs": "Yes",
-        "restecg": "Left ventricular hypertrophy", "thalach": 120, "exang": "Yes",
-        "oldpeak": 3.0, "slope": "Downsloping", "ca": "3",
-        "thal": "Fixed defect"
+    "sore throat": {
+        "assessment": "Often viral, but could be strep throat",
+        "recommendations": [
+            "Gargle with warm salt water 3-4 times daily",
+            "Drink warm liquids (tea with honey)",
+            "Use throat lozenges",
+            "Stay hydrated",
+            "Rest your voice"
+        ],
+        "warnings": [
+            "Difficulty swallowing or breathing",
+            "White patches on tonsils",
+            "Fever >38.5°C with throat pain"
+        ]
     }
 }
 
-def calculate_risk(inputs):
-    """Calculate heart disease risk based on clinical parameters"""
-    risk_factors = 0
-    total_possible = 13  # Total parameters being evaluated
-    risk_details = []
-    risk_params = []
-
-    if inputs['age'] > 60:
-        risk_factors += 1
-        risk_details.append(f"- Age: {inputs['age']} (Risk: 60+)")
-        risk_params.append('age')
-    if inputs['sex'] == 'Male':
-        risk_factors += 1
-        risk_details.append("- Sex: Male (Higher risk)")
-        risk_params.append('sex')
-    if inputs['cp'] == 'Asymptomatic':
-        risk_factors += 1
-        risk_details.append("- Chest Pain: Asymptomatic (Highest risk)")
-        risk_params.append('cp')
-    elif inputs['cp'] in ['Atypical angina', 'Non-anginal pain']:
-        risk_factors += 0.5
-        risk_details.append(f"- Chest Pain: {inputs['cp']} (Moderate risk)")
-    if inputs['trestbps'] >= 140:
-        risk_factors += 1
-        risk_details.append(f"- Blood Pressure: {inputs['trestbps']} mmHg (Stage 2 Hypertension)")
-        risk_params.append('trestbps')
-    elif inputs['trestbps'] >= 130:
-        risk_factors += 0.5
-        risk_details.append(f"- Blood Pressure: {inputs['trestbps']} mmHg (Elevated)")
-    if inputs['chol'] >= 240:
-        risk_factors += 1
-        risk_details.append(f"- Cholesterol: {inputs['chol']} mg/dL (High)")
-        risk_params.append('chol')
-    elif inputs['chol'] >= 200:
-        risk_factors += 0.5
-        risk_details.append(f"- Cholesterol: {inputs['chol']} mg/dL (Borderline High)")
-    if inputs['fbs'] == 'Yes':
-        risk_factors += 1
-        risk_details.append("- Fasting Blood Sugar > 120 mg/dL")
-        risk_params.append('fbs')
-    if inputs['restecg'] == 'Left ventricular hypertrophy':
-        risk_factors += 1
-        risk_details.append("- ECG: Left Ventricular Hypertrophy")
-        risk_params.append('restecg')
-    elif inputs['restecg'] == 'ST-T wave abnormality':
-        risk_factors += 0.5
-        risk_details.append("- ECG: ST-T Wave Abnormality")
-    if inputs['thalach'] < 120:
-        risk_factors += 1
-        risk_details.append(f"- Max Heart Rate: {inputs['thalach']} (Low)")
-        risk_params.append('thalach')
-    if inputs['exang'] == 'Yes':
-        risk_factors += 1
-        risk_details.append("- Exercise Induced Angina: Yes")
-        risk_params.append('exang')
-    if inputs['oldpeak'] >= 2.0:
-        risk_factors += 1
-        risk_details.append(f"- ST Depression: {inputs['oldpeak']} (High)")
-        risk_params.append('oldpeak')
-    elif inputs['oldpeak'] >= 1.0:
-        risk_factors += 0.5
-        risk_details.append(f"- ST Depression: {inputs['oldpeak']} (Moderate)")
-    if inputs['slope'] == 'Downsloping':
-        risk_factors += 1
-        risk_details.append("- ST Slope: Downsloping (Highest risk)")
-        risk_params.append('slope')
-    elif inputs['slope'] == 'Flat':
-        risk_factors += 0.5
-        risk_details.append("- ST Slope: Flat (Moderate risk)")
-    if inputs['ca'] == '3':
-        risk_factors += 1
-        risk_details.append("- Major Vessels: 3 (Highest risk)")
-        risk_params.append('ca')
-    elif inputs['ca'] in ['1', '2']:
-        risk_factors += 0.5
-        risk_details.append(f"- Major Vessels: {inputs['ca']} (Moderate risk)")
-    if inputs['thal'] == 'Fixed defect':
-        risk_factors += 1
-        risk_details.append("- Thalassemia: Fixed Defect")
-        risk_params.append('thal')
-    elif inputs['thal'] == 'Reversible defect':
-        risk_factors += 0.5
-        risk_details.append("- Thalassemia: Reversible Defect")
-
-    probability = min(risk_factors / total_possible, 0.99)  # Cap at 99%
-
-    if probability <= RISK_THRESHOLDS['Low']:
-        risk_level = 'Low'
-        color = "green"
-        icon = "✅"
-        advice = "Maintain your healthy lifestyle with regular check-ups."
-    elif probability <= RISK_THRESHOLDS['Moderate']:
-        risk_level = 'Moderate'
-        color = "orange"
-        icon = "⚠️"
-        advice = "Consider lifestyle changes and consult your doctor."
-    else:
-        risk_level = 'High'
-        color = "red"
-        icon = "❗"
-        advice = "Please consult a cardiologist immediately."
-
-    return risk_level, probability, color, icon, advice, risk_details, risk_params
-
-# App layout
-st.title("❤️ Heart Disease Risk Assessment")
-st.markdown("""
-This tool evaluates your heart disease risk based on clinical parameters and provides personalized insights.
-""")
-
-# Create tabs
-tab1, tab2, tab3 = st.tabs(["Risk Assessment", "Parameter Analysis", "Health Guidance"])
-
-with tab1:
-    with st.form("heart_form"):
-        st.subheader("Enter Your Health Parameters")
-        col1, col2, col3 = st.columns(3)
-        inputs = {}
-
-        # Column 1
-        with col1:
-            inputs["age"] = st.number_input(
-                feature_info["age"]["label"],
-                min_value=feature_info["age"]["min"],
-                max_value=feature_info["age"]["max"],
-                value=st.session_state.get("age", feature_info["age"]["value"]),
-                help=f"{feature_info['age']['desc']}. Healthy range: {feature_info['age']['healthy_range']}"
-            )
-            inputs["sex"] = st.selectbox(
-                feature_info["sex"]["label"],
-                feature_info["sex"]["options"],
-                index=feature_info["sex"]["options"].index(st.session_state.get("sex", feature_info["sex"]["options"][0])),
-                help=feature_info["sex"]["desc"]
-            )
-            inputs["cp"] = st.selectbox(
-                feature_info["cp"]["label"],
-                feature_info["cp"]["options"],
-                index=feature_info["cp"]["options"].index(st.session_state.get("cp", feature_info["cp"]["options"][0])),
-                help=f"{feature_info['cp']['desc']}. Healthy: {feature_info['cp']['healthy']}"
-            )
-            inputs["trestbps"] = st.number_input(
-                feature_info["trestbps"]["label"],
-                min_value=feature_info["trestbps"]["min"],
-                max_value=feature_info["trestbps"]["max"],
-                value=st.session_state.get("trestbps", feature_info["trestbps"]["value"]),
-                help=f"{feature_info['trestbps']['desc']}. Healthy: {feature_info['trestbps']['healthy_range']}"
-            )
-            inputs["chol"] = st.number_input(
-                feature_info["chol"]["label"],
-                min_value=feature_info["chol"]["min"],
-                max_value=feature_info["chol"]["max"],
-                value=st.session_state.get("chol", feature_info["chol"]["value"]),
-                help=f"{feature_info['chol']['desc']}. Healthy: {feature_info['chol']['healthy_range']}"
-            )
-
-        # Column 2
-        with col2:
-            inputs["fbs"] = st.selectbox(
-                feature_info["fbs"]["label"],
-                feature_info["fbs"]["options"],
-                index=feature_info["fbs"]["options"].index(st.session_state.get("fbs", feature_info["fbs"]["options"][0])),
-                help=f"{feature_info['fbs']['desc']}. Healthy: {feature_info['fbs']['healthy']}"
-            )
-            inputs["restecg"] = st.selectbox(
-                feature_info["restecg"]["label"],
-                feature_info["restecg"]["options"],
-                index=feature_info["restecg"]["options"].index(st.session_state.get("restecg", feature_info["restecg"]["options"][0])),
-                help=f"{feature_info['restecg']['desc']}. Healthy: {feature_info['restecg']['healthy']}"
-            )
-            inputs["thalach"] = st.number_input(
-                feature_info["thalach"]["label"],
-                min_value=feature_info["thalach"]["min"],
-                max_value=feature_info["thalach"]["max"],
-                value=st.session_state.get("thalach", feature_info["thalach"]["value"]),
-                help=f"{feature_info['thalach']['desc']}. Healthy: {feature_info['thalach']['healthy_range']}"
-            )
-            inputs["exang"] = st.selectbox(
-                feature_info["exang"]["label"],
-                feature_info["exang"]["options"],
-                index=feature_info["exang"]["options"].index(st.session_state.get("exang", feature_info["exang"]["options"][0])),
-                help=f"{feature_info['exang']['desc']}. Healthy: {feature_info['exang']['healthy']}"
-            )
-            inputs["oldpeak"] = st.number_input(
-                feature_info["oldpeak"]["label"],
-                min_value=feature_info["oldpeak"]["min"],
-                max_value=feature_info["oldpeak"]["max"],
-                value=st.session_state.get("oldpeak", feature_info["oldpeak"]["value"]),
-                step=feature_info["oldpeak"]["step"],
-                help=f"{feature_info['oldpeak']['desc']}. Healthy: {feature_info['oldpeak']['healthy_range']}"
-            )
-
-        # Column 3
-        with col3:
-            inputs["slope"] = st.selectbox(
-                feature_info["slope"]["label"],
-                feature_info["slope"]["options"],
-                index=feature_info["slope"]["options"].index(st.session_state.get("slope", feature_info["slope"]["options"][0])),
-                help=f"{feature_info['slope']['desc']}. Healthy: {feature_info['slope']['healthy']}"
-            )
-            inputs["ca"] = st.selectbox(
-                feature_info["ca"]["label"],
-                feature_info["ca"]["options"],
-                index=feature_info["ca"]["options"].index(st.session_state.get("ca", feature_info["ca"]["options"][0])),
-                help=f"{feature_info['ca']['desc']}. Healthy: {feature_info['ca']['healthy']}"
-            )
-            inputs["thal"] = st.selectbox(
-                feature_info["thal"]["label"],
-                feature_info["thal"]["options"],
-                index=feature_info["thal"]["options"].index(st.session_state.get("thal", feature_info["thal"]["options"][0])),
-                help=f"{feature_info['thal']['desc']}. Healthy: {feature_info['thal']['healthy']}"
-            )
-
-        submitted = st.form_submit_button("Assess My Heart Disease Risk")
-
-    if submitted:
-        risk_level, prediction_proba, color, icon, advice, risk_details, risk_params = calculate_risk(inputs)
-
-        st.subheader("Risk Assessment Results")
-        cols = st.columns(3)
-        with cols[1]:
-            st.metric("Heart Disease Risk", f"{prediction_proba * 100:.1f}%", f"{risk_level} Risk", delta_color="off")
-        risk_meter = st.progress(prediction_proba)
-        st.markdown(f"""
-        <div style="background-color:#f0f2f6;padding:15px;border-radius:10px">
-            <h4 style="color:{color};text-align:center">{icon} {risk_level} Risk Category</h4>
-            <p style="text-align:center">{advice}</p>
-        </div>
-        """, unsafe_allow_html=True)
-
-        st.subheader("Risk Factor Analysis")
-        total_factors = 13
-        risk_factors = len(risk_details)
-        st.write(f"**{risk_factors} out of {total_factors}** parameters show elevated risk:")
-        if risk_details:
-            st.markdown("\n".join(risk_details))
-        else:
-            st.success("All parameters are within healthy ranges!")
-
-        st.session_state.results = {
-            "risk_level": risk_level,
-            "probability": prediction_proba,
-            "risk_factors": risk_details,
-            "risk_params": risk_params,
-            "inputs": inputs
-        }
-
-with tab2:
-    st.header("Parameter Analysis")
-    if "results" not in st.session_state:
-        st.warning("Please complete the risk assessment first.")
-    else:
-        st.subheader("Health Parameter Radar Chart")
-        radar_params = ["age", "trestbps", "chol", "thalach", "oldpeak"]
-        radar_labels = [feature_info[p]["label"] for p in radar_params]
-        radar_values = [st.session_state.results["inputs"][p] for p in radar_params]
-
-        max_values = {
-            "age": 100,
-            "trestbps": 200,
-            "chol": 600,
-            "thalach": 220,
-            "oldpeak": 6.0
+def get_local_advice(symptom):
+    """Get structured medical advice from local knowledge base"""
+    symptom_lower = symptom.lower()
+    for key in MEDICAL_KNOWLEDGE:
+        if key in symptom_lower:
+            knowledge = MEDICAL_KNOWLEDGE[key]
+            response = "🔍 [Assessment]\n- " + knowledge["assessment"] + "\n\n"
+            response += "💡 [Self-Care Recommendations]\n" + "\n".join(["• " + item for item in knowledge["recommendations"]]) + "\n\n"
+            response += "⚠️ [When to Seek Medical Care]\n" + "\n".join(["• " + item for item in knowledge["warnings"]]) + "\n\n"
+            response += "📅 [Follow-up]\n• Re-evaluate in 2-3 days if not improving\n• See doctor if symptoms worsen or persist beyond 5 days"
+            return response
+    return None
+
+def query_mistral(prompt):
+    """Try API first, fallback to local knowledge"""
+    try:
+        headers = {"Authorization": f"Bearer {MISTRAL_API_KEY}", "Content-Type": "application/json"}
+        payload = {
+            "model": "mistral-tiny",
+            "messages": [{"role": "user", "content": prompt}],
+            "temperature": 0.7,
+            "max_tokens": 1024
         }
-        normalized_values = [v / max_values[p] for v, p in zip(radar_values, radar_params)]
-        angles = np.linspace(0, 2 * np.pi, len(radar_labels), endpoint=False).tolist()
-        angles += angles[:1]
-        normalized_values += normalized_values[:1]
-        fig, ax = plt.subplots(figsize=(6, 6), subplot_kw=dict(polar=True))
-        ax.fill(angles, normalized_values, color='red', alpha=0.25)
-        ax.plot(angles, normalized_values, color='red', linewidth=2)
-        ax.set_xticks(angles[:-1])
-        ax.set_xticklabels(radar_labels)
-        ax.set_yticklabels([])
-        ax.set_title("Your Health Parameters (Normalized)", pad=20)
-        st.pyplot(fig)
-
-        st.subheader("Parameter Comparison")
-        comparison_data = []
-        for param in ["age", "trestbps", "chol", "thalach", "oldpeak"]:
-            info = feature_info[param]
-            status = "⚠️ Out of Range" if param in st.session_state.results["risk_params"] else "✅ Within Range"
-            comparison_data.append({
-                "Parameter": info["label"],
-                "Your Value": st.session_state.results["inputs"][param],
-                "Healthy Range": info.get("healthy_range", info.get("healthy", "")),
-                "Status": status
-            })
-        st.table(pd.DataFrame(comparison_data))
-
-with tab3:
-    st.header("Personalized Health Guidance")
-    if "results" not in st.session_state:
-        st.warning("Please complete the risk assessment first.")
-    else:
-        risk_level = st.session_state.results["risk_level"]
-        
-        # Basic guidance
-        if risk_level == "Low":
-            st.success("""
-            **Your heart health looks good!**  
-            Maintain these healthy habits:
-            - Continue regular exercise (150 mins/week moderate activity)
-            - Eat a balanced diet rich in fruits, vegetables, and whole grains
-            - Get 7-9 hours of quality sleep nightly
-            - Manage stress through relaxation techniques
-            - Annual check-ups with your doctor
-            """)
-        elif risk_level == "Moderate":
-            st.warning("""
-            **Your heart health needs attention**  
-            Recommended actions:
-            - Increase physical activity (aim for 30 mins/day)
-            - Reduce sodium (<1500mg/day) and saturated fat intake
-            - Quit smoking if applicable
-            - Limit alcohol to 1 drink/day (women) or 2/day (men)
-            - Monitor blood pressure regularly
-            - Schedule a doctor's visit within 3 months
-            - Consider cholesterol screening
-            """)
-        else:
-            st.error("""
-            **Your heart health requires immediate attention**  
-            Critical next steps:
-            - Consult a cardiologist within 1 month
-            - Begin a supervised exercise program
-            - Strict dietary modifications (Mediterranean diet recommended)
-            - Medication may be needed (doctor will advise)
-            - Regular monitoring of blood pressure and cholesterol
-            - Consider cardiac rehabilitation program
-            - Emergency care if chest pain or shortness of breath occurs
-            """)
-        
-        # AI-Powered Recommendations
-        st.subheader("AI-Powered Recommendations")
-        if st.button("Get Personalized AI Advice"):
-            with st.spinner("Generating AI recommendations..."):
-                try:
-                    # Prepare the prompt for Mistral API
-                    prompt = f"""
-                    Provide detailed, personalized heart health recommendations for a patient with the following profile:
-                    - Age: {st.session_state.results["inputs"]["age"]}
-                    - Sex: {st.session_state.results["inputs"]["sex"]}
-                    - Risk Level: {risk_level}
-                    - Key Risk Factors: {', '.join(st.session_state.results["risk_params"])}
-                    
-                    Provide specific, actionable advice on:
-                    1. Lifestyle modifications
-                    2. Dietary recommendations
-                    3. Exercise suggestions
-                    4. Monitoring suggestions
-                    5. When to seek medical attention
-                    
-                    Format the response with clear headings and bullet points.
-                    """
-                    
-                    # Call Mistral API
-                    headers = {
-                        "Authorization": f"Bearer {MISTRAL_API_KEY}",
-                        "Content-Type": "application/json"
-                    }
-                    
-                    data = {
-                        "model": "mistral-tiny",
-                        "messages": [{"role": "user", "content": prompt}]
-                    }
-                    
-                    response = requests.post(
-                        "https://api.mistral.ai/v1/chat/completions",
-                        headers=headers,
-                        json=data
-                    )
-                    
-                    if response.status_code == 200:
-                        ai_response = response.json()["choices"][0]["message"]["content"]
-                        st.markdown("### AI Health Advisor Recommendations")
-                        st.write(ai_response)
-                    else:
-                        st.error(f"Failed to get AI recommendations. Status code: {response.status_code}")
-                
-                except Exception as e:
-                    st.error(f"An error occurred: {str(e)}")
-
-# Sidebar with sample profiles
-st.sidebar.header("Quick Start")
-profile = st.sidebar.selectbox("Load sample profile:", list(SAMPLE_PROFILES.keys()))
-if st.sidebar.button("Load Profile"):
-    for key in list(st.session_state.keys()):
-        if key in feature_info:
-            del st.session_state[key]
-    for key, value in SAMPLE_PROFILES[profile].items():
-        st.session_state[key] = value
-    st.rerun()
+        response = requests.post(MISTRAL_API_URL, headers=headers, json=payload, timeout=10)
+        response.raise_for_status()
+        return response.json()['choices'][0]['message']['content']
+    except:
+        return None
+
+def generate_response(name, message):
+    """Generate medical response with fallback"""
+    # First try local knowledge base
+    local_response = get_local_advice(message)
+    if local_response:
+        return f"Hello {name},\n\n{local_response}\n\nWishing you good health,\nDr. Alex"
+    
+    # Then try API
+    prompt = f"""As Dr. Alex, provide structured advice for {name} who reports: "{message}"
+
+    Format response with these sections:
+    
+    [Assessment] - Brief professional evaluation
+    [Self-Care Recommendations] - 3-5 specific actionable steps
+    [When to Seek Medical Care] - Clear warning signs
+    [Follow-up] - Monitoring advice
+    
+    Use bullet points, professional yet compassionate tone."""
+    
+    api_response = query_mistral(prompt)
+    if api_response:
+        return api_response
+    
+    # Final fallback
+    common_conditions = "\n".join([f"- {cond}" for cond in MEDICAL_KNOWLEDGE.keys()])
+    return f"""Hello {name},
+
+I'm currently unable to access detailed medical databases. For general advice:
+
+Common conditions I can advise on:
+{common_conditions}
+
+For immediate concerns:
+• Contact your local healthcare provider
+• In emergencies, call your local emergency number
+
+Please try asking about one of these common conditions or consult a healthcare professional.
+
+Best regards,
+Dr. Alex"""
+
+# Gradio Interface
+with gr.Blocks(title="Dr. Alex Medical Advisor", theme=gr.themes.Soft()) as demo:
+    gr.Markdown("# 🩺 Dr. Alex - General Health Advisor")
+    name = gr.Textbox(label="Your Name", placeholder="Enter your name")
+    
+    with gr.Tab("Health Consultation"):
+        gr.Markdown("## Describe your health concern")
+        msg = gr.Textbox(label="Symptoms/Question", lines=3)
+        submit_btn = gr.Button("Get Medical Advice")
+        output = gr.Textbox(label="Dr. Alex's Advice", lines=15, interactive=False)
+    
+    submit_btn.click(
+        fn=lambda name, msg: generate_response(name, msg),
+        inputs=[name, msg],
+        outputs=output
+    )
+    
+    gr.Markdown("### Common Health Topics")
+    with gr.Row():
+        for condition in MEDICAL_KNOWLEDGE:
+            gr.Button(condition.capitalize()).click(
+                fn=lambda name, cond=condition: generate_response(name, cond),
+                inputs=[name],
+                outputs=output
+            )
 
-# Disclaimer
-st.sidebar.markdown("""
-### Important Disclaimer
-This tool provides risk assessment only and is not a substitute for professional medical advice. 
-Always consult with a qualified healthcare provider for medical concerns.
-""")
+demo.launch()